Exploring Bank Stock Fundamentals – Part 2 – ROA

In this article, I’m going to examine in impact of return on assets (ROA) on bank stock returns. As I mentioned in my previous post exploring bank stock fundamentals, in order to understand why average excess returns decrease with increasing return on equity (ROE) we need to separate out ROA and leverage out from ROE. Given that ROA is a profitability measure, I expect that stocks with higher ROA would produce greater 1-year returns. Let’s run a backtest to find out.

Starting with my standard method of backtesting stock fundamentals on Portfolio123 that I described in an earlier article, I then just selected stocks in the banking subsector and changed my benchmark to the Financial Select Sector SPDR Fund (XLF). The average excess returns for each quintile are displayed in the chart below.

Bar chart of bank stock average excess returns by quintile of ROA (2000 – 2017)
Bank stock average excess returns by quintile of ROA (2000 – 2017)

While average excess returns increased from the 1st to the 2nd quintile, returns then dropped off for the top three quintiles as ranked by ROA. These results are not conclusive on their own. The full details of the backtest are in the table below.

Bar chart of the 18-year backtest of ROA for the banking sector
18-year backtest of ROA for the banking sector

The value of $10,000 invested on January 2, 2000 in the banking sector (equally weighted positions that are annually rebalanced) would be $28,960 at the end of 2017 for the lowest 20% of bank stocks ranked by ROA. Over that same time period, the value of $10,000 invested in the top 20% of bank stocks ranked by ROA would have been $34,480.  While this is higher than the results for the first quintile, I would have expected average excess returns to also have been higher for the 5th quintile versus the 1st quintile. I also note that total returns don’t increase evenly for each quintile. The 2nd quintile produced a total return of 315.51%, while the 5th quintile only produced a total return of 244.80% from 2000 to the end of 2017.

I did notice that maximum drawdown and maximum loss decreased as ROA increased over the five quintiles. However, maximum gain did not have a consistent trend as ROA increased.

While examining the ROA component of ROE didn’t really solve the unanticipated results from my earlier analysis of ROE, there is still the possibility that leverage is the key to describing why average excess returns seem to decline as return on equity increases when examining the banking sector.

2 thoughts on “Exploring Bank Stock Fundamentals – Part 2 – ROA

  • December 5, 2018 at 9:18 pm
    Permalink

    Looks like there is more going on there. What if you segment by book of assets / size and then run the backtest? We know from Fama that ME/BE (value) and ME (size) are very important in explaining returns… here you are looking at a different dimension altogether. If you combine both approaches I think you might get the ascending bar chart one would hope for.

    Leverage can’t explain why the roa chart doesn’t look right.

    Hope this helps…

  • December 6, 2018 at 9:36 am
    Permalink

    Thanks for the suggestion, Pedro. I’ll try your suggestion to segment the banks by size.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.